>> Occam code is posited to exploit structure, not merely be a compact
>> description or representation.

Michael> Much as I like Occam, that's a pretty strong claim for any
Michael> language designed primarily for ease of use by human
Michael> engineers, indeed for any 'substrate' without an integral
Michael> optimisation mechanism. 

My bad language. I didn't mean code written in the language Occam
(which I'd forgotten about, don't know anything about it).

I meant, code in some constrained language (for example, constrained
by being very concise) that solves so many different problems that
such constrained code wouldn't exist unless it was fundamentally
exploiting structure underlying the problem--
code exemplifying my generalized Occam's Razor hypothesis.

(I don't think I used the term "occam code" in the book. This is
another reason why offhand posts on the subject are strewing 
confusion...)

Many people argued that Lisp and
Michael> Prolog are/were 'posited to exploit structure' at the
Michael> language level, and while that worked well for a few early
Michael> projects (such as AM) I think the ultimate error of this
Michael> assumption was a significant contributing factor to the AI
Michael> winter. In our system we're using a language optimised as
Michael> much as possible for AI code generation (as opposed to human
Michael> use; not an uncommon strategy in AGI), somewhat like a cross
Michael> between Erlang, Occam and Flare (the language the SIAI was
Michael> trying to design back in 2001). Even still, almost all the
Michael> modelling power comes from the code pattern library and
Michael> associated mechanisms for reasoning about code, not the
Michael> language (which primarily makes the latter easier to write
Michael> and provides substrate support for various forms of more
Michael> sophisticated reflective reasoning).

Michael> Michael Wilson, Director of Research and Development,
Michael> Bitphase AI Ltd Web demos page: http://www.bitphase.com/apex

Michael> ----- This list is sponsored by AGIRI:
Michael> http://www.agiri.org/email To unsubscribe or change your
Michael> options, please go to:
Michael> http://v2.listbox.com/member/?list_id=303

-----
This list is sponsored by AGIRI: http://www.agiri.org/email
To unsubscribe or change your options, please go to:
http://v2.listbox.com/member/?list_id=303

Reply via email to